Publication number | US7436781 B2 |
Publication type | Grant |
Application number | US 11/143,253 |
Publication date | Oct 14, 2008 |
Filing date | Jun 2, 2005 |
Priority date | Jun 2, 2005 |
Fee status | Paid |
Also published as | US20060274676, WO2006132721A1 |
Publication number | 11143253, 143253, US 7436781 B2, US 7436781B2, US-B2-7436781, US7436781 B2, US7436781B2 |
Inventors | Feng Niu, Jian Huang, Spyros Kyperountas, Qicai Shi |
Original Assignee | Motorola, Inc. |
Export Citation | BiBTeX, EndNote, RefMan |
Patent Citations (8), Non-Patent Citations (2), Classifications (7), Legal Events (4) | |
External Links: USPTO, USPTO Assignment, Espacenet | |
The present invention relates generally to radiolocation and in particular, to a method and apparatus for determining the location of a node within a wireless communication system.
A variety of systems have been proposed that call for the ability to determine the location of an object within a wireless communication system. For example, in asset control, it is desirable to know the locations of objects (e.g., laptop computers, cell phones, . . . , etc.) within the confines of an office building. One technique utilized to locate objects is a power-based location technique. Such a technique relates the received signal strength (RSS) of a node to its distance. Although this technique can be very accurate in locating nodes within a communication system, power-based location techniques require a path loss model of the environment in order to perform inter-node range estimation. Thus, one has to carry out a measurement campaign to establish the relationship between the inter-node path loss and the inter-node distance, with path loss being a difference between a power a signal was received at and a power the signal was transmitted at. In the measurement campaign, the distances between many pairs of nodes have to be measured manually. When there are significant changes in the environment, the path loss model has to be re-established and thus measurement campaigns have to be carried out again. This makes any power-based location technique very cumbersome. Therefore, a need exists for a method and apparatus for locating a node within a wireless communication system that utilizes a power-based location technique, yet does not require the distances between pairs of nodes to be measured manually.
In order to address the above-mentioned need, a method and apparatus for determining the location of an object within a wireless communication system is provided herein. During operation, a path loss model for a node is generated based on path loss values and corresponding numbers of nodes on the neighbor list of the node. The path loss model is used to determine a relationship between path loss and distance. With this relationship established, distances to known-located nodes may be obtained by obtaining a path loss to the known-located node. From these distances, a node can then be located.
Because a path loss model is obtained based on a number of neighbors, there is no need to carry out a measurement campaign to establish the relationship between the inter-node path loss and the inter-node distance. Thus, there is no need to manually measure the distances between many pairs of nodes to establish a path loss model.
The present invention encompasses a method for locating a node within a wireless communication system. The method comprises the steps of determining a plurality of path loss values to a plurality of nodes, determining a path loss model based on a node density and the path loss to the plurality of nodes, and determining distances to known located nodes based on the path loss model. The node is then located based on distances to known located nodes.
The present invention additionally encompasses a method comprising the steps of determining a plurality of neighboring nodes, determining number of neighboring nodes having a path loss less than a particular neighboring node, determining a path loss model based on a node density and the number of nodes having the path loss less than the particular node, and determining distances to known-located nodes based on the path loss model. A location is then determined based on the distances to known-located nodes.
The present invention additionally encompasses an apparatus comprising a receiver receiving signals from a plurality of nodes, and logic circuitry analyzing the signals from the plurality of nodes to determine a number of nodes having a path loss less than a particular node, determine a path loss model based on a node density and a number of nodes having the path loss less than the particular node, determine distances to known-located nodes based on the path loss model, and determine a location based on the distances to known-located nodes.
Turning now to the drawings, wherein like numerals designate like components,
Circular objects, or nodes 104 (only one labeled) represent wireless devices, the locations of which are unknown and to be determined. Nodes 104 can include, for example, transceiver security tags attached to valuable assets such as lap top computers, laboratory equipment, or be embedded in wireless communication devices including cellular telephones. Rectangular objects 105 (only one labeled) represent reference nodes. The locations of nodes 105 are known, or can be easily and accurately determined to within some measurement accuracy (e.g., via physical measurement or via GPS).
It should be noted that although
It should be noted that in the preferred embodiment of the present invention each node 104 is capable of determining its own location based on a path-loss model and path-loss measurements to known-located nodes 105, however in an alternate embodiment of the present invention, centralized location-finding equipment 106 is provided to calculate locations. In the alternate embodiment of the present invention, nodes 104 will provide distance estimates to equipment 106. Location-finding equipment 106 will then use the provided distances to calculate locations for nodes 104.
As discussed above, although power-based location techniques can be very accurate in locating nodes within a communication system, such techniques require a path loss model of the environment in order to perform inter-node range estimation. Thus, one has to carry out a measurement campaign to establish the relationship between the inter-node path loss and the inter-node distance. In the measurement campaign, the distances between many pairs of nodes have to be measured manually. In order to address this issue, a path-loss model for a node 104 is determined based on path loss values and the corresponding numbers of neighbors of node 104. The path-loss model determination done in this manner does not require inter-node distance measurement.
Path Loss Model Determination
Although many different equations relating path loss to distance may be used, for ease of explanation, the following description will use a generic RSSI path-loss model. In such a generic path loss model, a path loss between two nodes separated by a distance r includes a range independent term, a range dependent term, and a random error term as follows:
PL(in dB)=PL _{0}(in dB)+F(r)(in dB)+ε. (1)
PL is the path-loss between two nodes and is calculated by dividing the transmitted power by the received power and converting the result to dB. To convert to dB, one takes a 10-based logarithm of the result and multiplies the result by 10. PL_{0 }is the path-loss at some constant distance r_{0 }(r_{0}=1 m here for convenience) and is a constant to be estimated. ε is the random error representing mainly the error due to the propagation environment or shadowing effect. We are interested in estimating the deterministic portion of the path-loss model in (1), i.e., the range independent term PL_{0 }and range dependent term F(r).
Equation (1) can be interpreted in different ways. One way is to stand at a node 201 and look out from the node (see
If the RSSI (transmit and receive power) information is available in addition to the neighbor list, each node can generate its own path-loss model based on this information without carrying out a measurement campaign consisting of many range (r) measurements.
The basic concept relates the radius r from a node to N, the number of neighbors of the said node within the radius r. For example, in a 2D area and assuming no shadowing effect, the area within the radius r of a node is πr^{2}. If the density of nodes is ρ, the number of nodes within the radius r is:
N=ρπr^{2}, (2a)
or,
r=(N/πρ)^{1/2}. (2b)
An example is shown in the
Once the relationship between N and r, i.e., r(N), is established, equation (1) can be re-written as
PL(in dB)=PL _{0}(in dB)+F(r(N))(in dB)+ε. (1a)
N can be obtained directly from the neighbor list with the RSSI information. To do this, each node ranks its neighbors according to the path-loss information and then a 1-to-1 correspondence between the path-loss PL and the number of neighbors N with their path-loss no greater than PL can be established.
As is evident, to establish the relationship (1) between PL and r, a lot of manual measurements of distance r are required, while to establish the relationship between PL and N in (1a), no extra measurements are needed other than the available neighbor list and RSSI information. To show how this works, let's assume the following commonly used F(r),
F(r)=10n log(r). (3)
Where n is the path loss index.
Equation (1) becomes
PL(in dB)=PL _{0}(in dB)+10n log(r)+ε. (1b)
From equations (2b) and (3), we have
F(r)=5n log(N)−5n log(πρ)+ε_{1}. (4)
Combining (1) and (4), we have
PL(dB)=PL _{0}(dB)+5n log(N)−5n log(πρ)+ε+ε_{1}. (5)
ε_{1 }is the error due to the replacement of r in F(r) by r(N), including the boundary effect and quantization error (N taking integer numbers only). These errors will show up in the actual implementations and their effect will be discussed later. If we have the neighbor list for a node and the RSSI information, we can arrange the neighbor list of node according to the path-loss or RSSI, say from small path loss to large path loss (closest to farthest). Equation (5) relates a given path-loss PL and the corresponding N, number of nodes whose path loss to the said node is no greater than the path loss PL on the left-hand side of equation (5). Equation (5) can be re-written as
y=b _{0} +b _{1} x+ε, (6)
y=PL(dB), x=log(N), b _{0} =PL _{0}−5n log(πρ), b _{1}=5n. (6a)
One can estimate both parameters b_{0 }and b_{1 }as follows,
The summation is over all statistical samples to be used for the parameter estimation. N_{max }is the total number of samples used. For example, with reference to
N_{max}=2;
Sample 1 at r _{1} : N _{1}=8, y _{1}=60 dB, x _{1}=log(N _{1})
Sample 2 at r _{2} : N _{2}=34, y _{2}=70 dB, x _{2}=log(N _{2}).
From (6) and (6a), we can obtain the estimation of n and PL_{0 }as follows
n=b _{1}/5, (8a)
PL _{0} =b _{0} +b _{1 }log(πρ). (8b)
If we know the node density ρ, we can obtain both PL_{0 }and n. If, instead of a 2D area, we have a 1D or a 3D space, (7a) and (7b) still hold but (8a) has to be modified to
n=m b _{1}/10, (9)
m=1, 2, 3 for 1D, 2D, and 3D, respectively. The relationship between N and r for 1D is,
N=ρπr, or, r=N/ρπ. (10)
And for 3D,
N=(4/3)ρπr ^{3}, or, r=(3N/4ρπ)^{1/3}. (11)
The node density functions ρ in (10), (2a), and (11) should use the proper units for 1D (nodes/m), 2D (nodes/M^{2}), and 3D (nodes/M^{3}), respectively.
For other forms of F(r), similar equations as (5-9) can be obtained to estimate all the parameters in equation (1). If the node density is known, we can obtain the range independent term PL_{0 }in equation (1).
Due to the shadowing effect, the quality of the data will deteriorate for nodes close to the boundary of the system coverage area. For example, if a node has a range of 10 m and is 15 m away from the boundary, the path loss index n in equation (1a) is 2, and the shadowing effect has a standard deviation of 6 dB, the node will be able to communicate with nodes beyond 15 m and thus the estimated path loss index will depend on nodes beyond 15 m. Since the said node is 15 m away from the boundary, the relationship between N and r (from (10), (2a), and (11) for 1D, 2D, and 3D, respectively) will not hold, as there will be no more nodes beyond the boundary. It is then preferred to use the data depending only on nodes within the boundary, which may count for only certain percentage of data points with path loss smaller than a certain quantity. This percentage number may range from a few percent to 100% and will depend on the index n, shadowing effect of the environment, and the closeness of the node to the system boundary. The smaller n, stronger shadowing, and closer to the boundary require lower percentage of nodes to be counted or need to include the nodes with smaller path loss.
In replacing r in (1b) with N as in (5), we have introduced a quantization error included in ε_{1}. This error comes essentially due to the fact that we represent a continuous distance variable r by a discrete number N as in equations (10), (2b), and (11) for 1D, 2D, and 3D, respectively. This error can be reduced by using N−1 rather than N in (2b), (10), and (11), where N is the number of neighbors within the radius r.
The shadowing may also have an effect of increasing the density of nodes, i.e., a node may have more neighbors when the shadowing is strong than when the shadowing is weak.
Implementation
Once the distances to known-located nodes are known, logic circuitry 305 locates itself. It should be noted that if centralized LFE 106 is being utilized to determine locations, logic circuitry 305 would then estimate a path loss model, determine distances to know-located nodes, and report the distances to LFE 106, where a location estimate will be made.
Once the above steps are taken, then logic circuitry can determine {circumflex over (b)}_{1 }and {circumflex over (b)}_{0}. Logic circuitry then determines a value for n from {circumflex over (b)}_{1}, and PL_{0 }from {circumflex over (b)}_{0}. Once n and PL_{0 }are obtained, a distance to each neighboring node can be estimated from equation 1b and a relationship between path loss and distance (i.e., the path loss model) is obtained. As is evident, the distance to each neighboring node is not obtained physically, but is estimated based on a node density (e.g., nodes per square meter, or nodes per cubic meter) and a number of nodes having a path loss less than a particular value.
Continuing, once a path loss model has been obtained, the logic flow continues to step 403 where distances to known located nodes are obtained based on the path loss model. In particular, logic circuitry 305 obtains a path loss for each known-located node, and this path loss is translated to a distance based on the path loss model. Once these distances have been obtained, the logic flow continues to step 405 where the distances can either be provided to centralized equipment for location (via transceiver 304), or logic circuitry 305 can locate the node.
While the invention has been particularly shown and described with reference to a particular embodiment, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention. For example, the above discussion was provided utilizing a particular path loss model shown in equation 1a. It should be evident to those of ordinary skill in the art that any path loss model may be utilized. Regardless of the path loss model used, node density, path loss values and corresponding numbers of neighbors of a node will be used to determine a distance of the said node to each of its neighbors. It is intended that such changes come within the scope of the following claims.
Cited Patent | Filing date | Publication date | Applicant | Title |
---|---|---|---|---|
US5613205 | Mar 31, 1995 | Mar 18, 1997 | Telefonaktiebolaget Lm Ericsson | System and method of locating a mobile terminal within the service area of a cellular telecommunication system |
US6140964 | Mar 18, 1997 | Oct 31, 2000 | Matsushita Electric Industrial Co., Ltd. | Wireless communication system and method and system for detection of position of radio mobile station |
US6990428 * | Jul 28, 2003 | Jan 24, 2006 | Cisco Technology, Inc. | Radiolocation using path loss data |
US7116988 * | Mar 16, 2004 | Oct 3, 2006 | Airespace, Inc. | Location of wireless nodes using signal strength weighting metric |
US7205938 * | Mar 5, 2004 | Apr 17, 2007 | Airespace, Inc. | Wireless node location mechanism responsive to observed propagation characteristics of wireless network infrastructure signals |
US7286829 * | Nov 17, 2004 | Oct 23, 2007 | Electronics And Telecommunications Research Institute | Base station selecting method in wireless network |
US7286835 * | Sep 10, 2004 | Oct 23, 2007 | Airespace, Inc. | Enhanced wireless node location using differential signal strength metric |
US20040203841 | Mar 13, 2002 | Oct 14, 2004 | Telefonaktiebolaget L M Ericsson (Publ) | Determination of mobile station location |
Reference | ||
---|---|---|
1 | Ghassemzadeh, S.S.; Jana, R.; Rice, C.W.; Turin, W.; Tarokh, V.: 'A statistical path loss model for in-home UWB channels', Ultra Wideband Systems and Technologies, 2002. Digest of Papers. 2002 IEEE Conference on , May 21-23, 2002, pp. 59-64. | |
2 | Patwari, N.; O'Dea, R.J.; Yanwei Wang: 'Relative location in wireless networks,'IEEE VTC'01, pp. 1149-1153. |
U.S. Classification | 370/252, 370/338 |
International Classification | G01R31/08 |
Cooperative Classification | G01S5/0289, H04B17/27 |
European Classification | G01S5/02R1, H04B17/00B6 |
Date | Code | Event | Description |
---|---|---|---|
Jun 2, 2005 | AS | Assignment | Owner name: MOTOROLA, INC., ILLINOIS Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:NLU, FENG;KYPEROUNTAS, SPYROS;SHI, QICAI;REEL/FRAME:016659/0138 Effective date: 20050531 |
Aug 4, 2009 | CC | Certificate of correction | |
Apr 6, 2011 | AS | Assignment | Owner name: MOTOROLA SOLUTIONS, INC., ILLINOIS Free format text: CHANGE OF NAME;ASSIGNOR:MOTOROLA, INC;REEL/FRAME:026081/0001 Effective date: 20110104 |
Mar 23, 2012 | FPAY | Fee payment | Year of fee payment: 4 |